Folashade Uba - Speaker Deck

Slide 1

Slide 1 text

Analysing Customer Insight, Unleashing SEO Potentials From Support Queries. Folashade Uba AP MOLLER - MAERSK @folashadeuba Speakerdeck.com/folashadeuba

Slide 2

Slide 2 text

No content

Slide 3

Slide 3 text

No content

Slide 4

Slide 4 text

Contact Form Email Calls

Slide 5

Slide 5 text

How to Leverage Customer Insights in Our SEO EFFORTS

Slide 6

Slide 6 text

“It’s important to remember that your website is not about YOU. It’s about YOUR VISITORS and their NEEDS” Derek Halpern

Slide 7

Slide 7 text

VALUE SALES CONVERSION

Slide 8

Slide 8 text

Why should we CARE about this as SEOs?

Slide 9

Slide 9 text

Understanding Customer Needs and Pain Points

Slide 10

Slide 10 text

Understanding Customer Needs and Pain Points Keyword and Topic Discovery

Slide 11

Slide 11 text

Understanding Customer Needs and Pain Points Keyword and Topic Discovery Improving Search Intent Alignment

Slide 12

Slide 12 text

Understanding Customer Needs and Pain Points Keyword and Topic Discovery Improving Search Intent Alignment Reduce bounce rate and increase conversions

Slide 13

Slide 13 text

Understanding Customer Needs and Pain Points Keyword and Topic Discovery Improving Search Intent Alignment Reduce bounce rate and increase conversions

Slide 14

Slide 14 text

SEO used to be a race to the top of the leaderboard

Slide 15

Slide 15 text

Now… it's about building a user-friendly map that guides customers directly to the solutions they need.

Slide 16

Slide 16 text

Now…We have all these customer feedback WHAT’S NEXT?

Slide 17

Slide 17 text

Large Dataset. Machine Learning Analysis. Case study: Qatar airways Customer review dataset gotten from Kaggle.com

Slide 18

Slide 18 text

Customer Pain Points Analysis(We are looking out for ….)

Slide 19

Slide 19 text

Customer Pain Points Analysis(We are looking out for ….) Uncover Search Intent

Slide 20

Slide 20 text

Customer Pain Points Analysis(We are looking out for ….) Uncover Search Intent Discover Themes and Patterns

Slide 21

Slide 21 text

Sample Data frame

Slide 22

Slide 22 text

Sentiment Analysis

Slide 23

Slide 23 text

Negative- 270 Positive- 2091 Neutral- 8 Reviews

Slide 24

Slide 24 text

Focus on the Negatives

Slide 25

Slide 25 text

Part-of-speech (POS) tagging What is POS and how is it to be used in the analysis? • Noun • Adverb • Adjective • Verbs Lemma

Slide 26

Slide 26 text

Negative Reviews (Adverbs and Adjectives) 114 110 102 80 63 61 55 53 45 39 39 36 0 20 40 60 80 100 120 when very so only then again even back just however never later Adverb Lemmas Contact Form Data 72 51 44 39 38 36 32 31 27 26 26 0 10 20 30 40 50 60 70 80 next other good more first last old poor full long uncomfortable Adjective Lemmas Contact Form Data

Slide 27

Slide 27 text

Negative Reviews (Verbs and Nouns) 2228 992 991 852 650 642 624 509 499 487 457 350 0 500 1000 1500 2000 2500 have fly get be check make take give serve go use offer Verbs Lemma Contact Forms Data 4391 2176 2027 1580 1552 1171 1109 1072 1014 891 879 864 0 1000 2000 3000 4000 5000 flight seat service time food staff class hour crew business cabin airline Noun Lemmas Contact Form Data

Slide 28

Slide 28 text

Sentiment analysis on Lemma

Slide 29

Slide 29 text

Negative Lemma

Slide 30

Slide 30 text

Words associated with these Negative lemma

Slide 31

Slide 31 text

Words associated with these Negative lemma

Slide 32

Slide 32 text

Filtering into the Dataset

Slide 33

Slide 33 text

Bring Back the data frame and perform POS * Focus on things, not strings.

Slide 34

Slide 34 text

Top Nouns(100)

Slide 35

Slide 35 text

Topic Modeling

Slide 36

Slide 36 text

Top 100 Nouns Semantically Categorized

Slide 37

Slide 37 text

• Airplane • Comfort • Flight • Meal • Time • Trip Themes 5 7 10 14 17 47 0 5 10 15 20 25 30 35 40 45 50 Trip Time Meal Flight Comfort Airplane number_unique _lemma Theme Number of Unique Lemmas per Cluster

Slide 38

Slide 38 text

No content

Slide 39

Slide 39 text

SYNTAX SHOWING USER’S INTENT VERBS NOUNS +

Slide 40

Slide 40 text

12 12 13 14 19 49 56 62 116 0 20 40 60 80 100 120 140 have time cancel flight use Qatar cdg doh fly experience fly business take care fly Qatar connect flight Most Common Action Pairs(Lemmatised Bigrams)

Slide 41

Slide 41 text

Why use Machine learning and not AI?

Slide 42

Slide 42 text

What do we then do with these INSIGHTS?

Slide 43

Slide 43 text

Develop new content pieces to fill these gaps

Slide 44

Slide 44 text

Optimize your content to address these needs

Slide 45

Slide 45 text

Incorporation of keywords into existing content

Slide 46

Slide 46 text

Leverage Social Media and other Channels

Slide 47

Slide 47 text

Utilise Structured Data

Slide 48

Slide 48 text

Enhance Local SEO

Slide 49

Slide 49 text

Content isn’t KING; it's the CUSTOMER who’s the KING KEY TAKEAWAY

Slide 50

Slide 50 text

Designing CONTENT tailored to their NEEDS and PROBLEMS will guide them smoothly toward CONVERSION. KEY TAKEAWAY

Slide 51

Slide 51 text

Let's Connect on LinkedIn!